基於大型語言模型應用指示詞打造無程式碼對話系統平台 - 以聊故事機器人為例

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：129

、訪客IP：3.17.175.239

姓名

林子平(Tzu-Ping Lin) 查詢紙本館藏

畢業系所

資訊工程學系

論文名稱

基於大型語言模型應用指示詞打造無程式碼對話系統平台 - 以聊故事機器人為例
(Prompts are All your Need to Construct a Dialogue System - A Case Study on Story Chatbots)

相關論文

★ 行程邀約郵件的辨識與不規則時間擷取之研究	★ NCUFree校園無線網路平台設計及應用服務開發
★ 網際網路半結構性資料擷取系統之設計與實作	★ 非簡單瀏覽路徑之探勘與應用
★ 遞增資料關聯式規則探勘之改進	★ 應用卡方獨立性檢定於關連式分類問題
★ 中文資料擷取系統之設計與研究	★ 非數值型資料視覺化與兼具主客觀的分群
★ 關聯性字組在文件摘要上的探討	★ 淨化網頁：網頁區塊化以及資料區域擷取
★ 問題答覆系統使用語句分類排序方式之設計與研究	★ 時序資料庫中緊密頻繁連續事件型樣之有效探勘
★ 星狀座標之軸排列於群聚視覺化之應用	★ 由瀏覽歷程自動產生網頁抓取程式之研究
★ 動態網頁之樣版與資料分析研究	★ 同性質網頁資料整合之自動化研究

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

過去的教育類型對話代理人存在兩個主要難點。首先，在語言理解能力有限的情況下，往往只能採取主導式的問答模式。這種模式下，對話代理人通常只能提供選項供學生選擇回答，而即使允許學生自由回答，回應通常也僅限於樣板式的回覆。這嚴重限制了對話系統的互動性和個性化，對教學成效造成了影響。其次，針對不同程度、背景的學生所採取的教學策略和方法通常需要有相對應的教學理論支持。然而，開發這樣的系統非常耗時且費工，往往僅能實作少數教學策略。
為了解決這個問題，我們建立了一個名為Educational Agent Crafting Tool（簡稱EduACT）的教育對話代理人創造平台。這個平台利用強大的語言模型ChatGPT，幫助教育工作者創造教學助教。與被動等待學生發問不同，我們採用了自然語言指示（Prompt）引導對話代理人，提供主動對話代理人類型設定。這使得教育工作者可以根據不同的教學內容和對象快速建立多個對話代理人，從而提供學生的一對一教學的可能。EduACT 平台的核心在於解構教學者與學生之間的對話方式，使其能夠以自然的方式與學生互動，回答問題、提供資訊並執行特定任務，從而增加教學的互動性。為了讓沒有程式背景的教育工作者容易使用，我們提出了一個模組化架構方法。創作者可以為對話代理人設計多個模組，每個模組負責不同的動作任務。這樣的設計使得對話過程比直接使用ChatGPT 更加主動和可控，處理多輪對話時回應更加一致合理，避免對話代理人提供不適當的回答，或是被學生引導至與課程無關的話題。
我們使用EduACT 平台設計了三個教師類型的對話代理人「魚姊姊」家族，以及三個學生類型的「熊小弟」們。我們讓兩者進行對話，並且顯示了它們能夠成功扮演各自角色，並且使用設定的任務模組進行完整的對話。總結來說，EduACT 平台提供了有教育專業背景但是沒有程式基礎的教育工作者直觀且容易上手的界面，讓他們能夠建立多元且高互動性的對話代理人，以確保對話代理人能夠更好地融入教學場景，提供更有效的教學輔助和學習體驗。

摘要(英)

In the past, educational conversational agents faced two main challenges. Firstly, their limited language understanding resulted in a predominantly directive question-and-answer format. This constrained interactivity and personalization, affecting teaching effectiveness. Secondly, creating tailored teaching strategies for students with different levels and backgrounds was time-consuming and challenging.
To tackle these issues, we developed Educational Agent Crafting Tool (EduACT), a platform that utilizes ChatGPT to help educators create teaching assistants. Instead of passively waiting for questions, EduACT uses natural language prompts to guide active conversations, enabling educators to quickly create multiple agents for personalized one-on-one instruction.EduACT deconstructs dialogue between educators and students, allowing natural interactions for answering questions, providing information, and executing tasks, thus increasing teaching interactivity. To simplify usage for educators without programming knowledge, we introduced a modular approach. This ensures more controlled and consistent responses during multi-turn conversations, avoiding irrelevant topics.
Using EduACT, we designed three teacher types “Fish Sisters” and three student types “Bear Brothers” for successful role portrayal and comprehensive conversations. Overall, EduACT empowers educators to create diverse, interactive conversational agents, enhancing teaching assistance and learning experiences.

關鍵字(中)

★ 對話系統
★ 對話代理人
★ 指示學習
★ 無程式碼人工智慧

關鍵字(英)

★ Dialogue System
★ Conversational Agent
★ Instruction Learning
★ No-Code Artificial Intelligence

論文目次

摘要 i
Abstract ii
目錄 iii
圖目錄 v
表目錄 vi
一、緒論 1
二、相關研究 5
2-1 聊故事機器人 5
2-2 對話代理人建置系統平台 7
三、研究背景 8
3-1 大型語言模型 8
3-2 指示微調 9
四、方法 11
4-1 對話代理人架構 12
4-2 對話代理人的建立 13
4-2-1 基本資訊 14
4-2-2 任務模組 14
4-2-3 動作策略 15
4-3 使用對話代理人 15
五、案例研究—聊故事機器人 16
5-1 老師類型對話代理人 16
5-1-1 上傳資料集 16
5-1-2 魚姊姊任務模組與使用規則 17
5-1-3 不同教學任務類型的魚姊姊任務模組與使用規則 18
5-2 學生類型對話代理人 19
5-2-1 熊小弟任務模組與使用規則 19
5-2-2 不同學生個性比較 20
5-3 對話範例 21
六、實驗 24
6-1 評估ChatGPT 與Chinese Alpaca 生成的對話 24
6-1-1 人工評估 25
6-2 系統效能測試 25
七、結論與未來工作 27
參考文獻 29
附錄一 33
A-1 EduACT 網頁畫面 33
A-2 資料表鍵值 33
A-3 對話代理人指示詞用法 35
A-3-1 任務選擇模組 36
A-3-2 對話生成模組 38

參考文獻

[1] character.ai. https://beta.character.ai/.
[2] Repika. https://replika.com/.
[3] Sebastian Wollny, Jan Schneider, Daniele Di Mitri, Joshua Weidlich, Marc Rittberger, and Hendrik Drachsler. Are we there yet?-a systematic literature review on chatbots in education. Frontiers in artificial intelligence, 4:654924, 2021.
[4] Mohammad Amin Kuhail, Nazik Alturki, Salwa Alramlawi, and Kholood Alhejori. Interacting with educational chatbots: A systematic review. Education and Information Technologies, 28(1):973–1018, 2023.
[5] José Quiroga Pérez, Thanasis Daradoumis, and Joan Manuel Marquès Puig. Rediscovering the use of chatbots in education: A systematic literature review. Computer Applications in Engineering Education, 28(6):1549–1565, 2020.
[6] Marina Umaschi Bers and Justine Cassell. Interactive storytelling systems for children: Using technology to explore language and identity. Journal of Interactive Learning Research, 9:183–215, 1998.
[7] Iulian V Serban, Chinnadhurai Sankar, Mathieu Germain, Saizheng Zhang, Zhouhan Lin, Sandeep Subramanian, Taesup Kim, Michael Pieper, Sarath Chandar, Nan Rosemary Ke, et al. A deep reinforcement learning chatbot. arXiv preprint arXiv:1709.02349, 2017.
[8] 陳臆玄. 應用強化式學習於多面向對話回應模組之研究. 碩士論文, 國立中央大學, 2022.
[9] Jason Wei, Maarten Bosma, Vincent Y Zhao, Kelvin Guu, Adams Wei Yu, Brian Lester, Nan Du, Andrew M Dai, and Quoc V Le. Finetuned language models are zero-shot learners. arXiv preprint arXiv:2109.01652, 2021. [10] Prakhar Gupta, Cathy Jiao, Yi-Ting Yeh, Shikib Mehri, Maxine Eskenazi, and Jeffrey Bigham. InstructDial: Improving zero and few-shot generalization in dialogue through instruction tuning. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 505–525, Abu Dhabi, United Arab Emirates, December 2022. Association for Computational Linguistics.
[11] Andy Zeng, Maria Attarian, brian ichter, Krzysztof Marcin Choromanski, Adrian Wong, Stefan Welker, Federico Tombari, Aveek Purohit, Michael S Ryoo, Vikas Sindhwani, Johnny Lee, Vincent Vanhoucke, and Pete Florence. Socratic models: Composing zero-shot multimodal reasoning with language. In The Eleventh International Conference on Learning Representations, 2023.
[12] Edward Y. Chang. Prompting large language models with the socratic method. In 2023 IEEE 13th Annual Computing and Communication Workshop and Conference (CCWC), pages 0351–0360, 2023.
[13] Chen-Chung Liu, Mo-Gang Liao, Chia-Hui Chang, and Hung-Ming Lin. An analysis of children’interaction with an ai chatbot and its impact on their interest in reading. Computers & Education, 189:104576, 2022.
[14] Jeya Amantha Kumar. Educational chatbots for project-based learning: investigating learning outcomes for a team-based design course. International journal of educational technology in higher education, 18(1):1–28, 2021.
[15] Khanmigo. https://www.khanacademy.org/khan-labs.
[16] OpenAI. Gpt-4 technical report. ArXiv, abs/2303.08774, 2023.
[17] Zheng Zhang, Ying Xu, Yanhao Wang, Bingsheng Yao, Daniel Ritchie, Tongshuang Wu, Mo Yu, Dakuo Wang, and Toby Jia-Jun Li. Storybuddy: A human-ai collaborative chatbot for parent-child interactive storytelling with flexible parental involvement. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, pages 1–21, 2022.
[18] Roger Greenaway. Playback: A guide to reviewing activities. In Duke of Edinburgh’s Award. association with Endeavour Scotland, 1993.
[19] Tzu-Hsien Huang and Chia-Hui Chang. Improving response diversity through commonsense-aware empathetic response generation. In Proceedings of the 34th Conference on Computational Linguistics and Speech Processing (ROCLING 2022), pages 299–306, 2022.
[20] Eleni Adamopoulou and Lefteris Moussiades. An overview of chatbot technology. In IFIP international conference on artificial intelligence applications and innovations, pages 373–383. Springer, 2020.
[21] Romal Thoppilan, Daniel De Freitas, Jamie Hall, Noam Shazeer, Apoorv Kulshreshtha, Heng-Tze Cheng, Alicia Jin, Taylor Bos, Leslie Baker, Yu Du, et al. Lamda: Language models for dialog applications. arXiv preprint arXiv:2201.08239, 2022.
[22] Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35:27730– 27744, 2022.
[23] Yiming Cui, Ziqing Yang, and Xin Yao. Efficient and effective text encoding for chinese llama and alpaca. arXiv preprint arXiv:2304.08177, 2023.
[24] Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971, 2023.
[25] Rohan Taori, Ishaan Gulrajani, Tianyi Zhang, Yann Dubois, Xuechen Li, Carlos Guestrin, Percy Liang, and Tatsunori B. Hashimoto. Stanford alpaca: An instruction-following llama model. https://github.com/tatsu-lab/stanford_alpaca, 2023.
[26] Yizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu, Noah A. Smith, Daniel Khashabi, and Hannaneh Hajishirzi. Self-instruct: Aligning language model with self generated instructions, 2022.
[27] Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. Language models are few-shot learners. Advances in neural information processing systems, 33:1877–1901, 2020.
[28] Renze Lou, Kai Zhang, and Wenpeng Yin. Is prompt all you need? no. a comprehensive and broader view of instruction learning. arXiv preprint arXiv:2303.10475, 2023.
[29] Ying Xu, Dakuo Wang, Mo Yu, Daniel Ritchie, Bingsheng Yao, Tongshuang Wu, Zheng Zhang, Toby Li, Nora Bradford, Branda Sun, Tran Hoang, YisiSang, Yufang Hou, Xiaojuan Ma, Diyi Yang, Nanyun Peng, Zhou Yu, and Mark Warschauer. Fantastic questions and where to find them: FairytaleQA – an authentic dataset for narrative comprehension. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 447–460, Dublin, Ireland, May 2022. Association for Computational Linguistics.
[30] Anaïs Tack and Chris Piech. The ai teacher test: Measuring the pedagogical ability of blender and gpt-3 in educational dialogues. arXiv preprint arXiv:2205.07540, 2022.
[31] Benjamin S Bloom. The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring. Educational researcher, 13(6):4–16, 1984.
[32] Weijiao Huang, Khe Foon Hew, and Luke K Fryer. Chatbots for language learning—are they really useful? a systematic review of chatbot-supported language learning. Journal of Computer Assisted Learning, 38(1):237–257, 2022.

指導教授

張嘉惠(Chia-Hui Chang)

審核日期

2023-7-28

推文